qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Peter Maydell <peter.maydell@linaro.org>
To: qemu-devel@nongnu.org
Cc: "Justin M. Forbes" <jmforbes@linuxtx.org>
Subject: [Qemu-devel] [PATCH 2/4] target-arm: Support v6 barriers in linux-user mode
Date: Wed, 27 Jul 2011 14:47:54 +0100	[thread overview]
Message-ID: <1311774476-23321-3-git-send-email-peter.maydell@linaro.org> (raw)
In-Reply-To: <1311774476-23321-1-git-send-email-peter.maydell@linaro.org>

ARMv6 implemented various operations as special cases of cp15 accesses
which are true instructions in v7; this includes barriers (DMB, DSB, ISB).
Catch this special case at translate time, so that it works in linux-user
mode (which doesn't provide a functional get_cp15 helper) as well as
system mode.

Includes minor cleanup of the existing cases (single switch statement,
and doing the "OK in user mode?" test explicitly rather than hiding it in
cp15_user_ok()).

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
 target-arm/translate.c |   51 +++++++++++++++++++++++++++++++----------------
 1 files changed, 33 insertions(+), 18 deletions(-)

diff --git a/target-arm/translate.c b/target-arm/translate.c
index 34d5e6e..c7961b8 100644
--- a/target-arm/translate.c
+++ b/target-arm/translate.c
@@ -2498,12 +2498,6 @@ static int cp15_user_ok(CPUState *env, uint32_t insn)
         if (op == 2 || (op == 3 && (insn & ARM_CP_RW_BIT)))
             return 1;
     }
-    if (cpn == 7) {
-        /* ISB, DSB, DMB.  */
-        if ((cpm == 5 && op == 4)
-                || (cpm == 10 && (op == 4 || op == 5)))
-            return 1;
-    }
     return 0;
 }
 
@@ -2579,39 +2573,60 @@ static int disas_cp15_insn(CPUState *env, DisasContext *s, uint32_t insn)
         /* cdp */
         return 1;
     }
-    if (IS_USER(s) && !cp15_user_ok(env, insn)) {
-        return 1;
-    }
-
-    /* Pre-v7 versions of the architecture implemented WFI via coprocessor
-     * instructions rather than a separate instruction.
+    /* We special case a number of cp15 instructions which were used
+     * for things which are real instructions in ARMv7. This allows
+     * them to work in linux-user mode which doesn't provide functional
+     * get_cp15/set_cp15 helpers, and is more efficient anyway.
      */
-    if ((insn & 0x0fff0fff) == 0x0e070f90) {
+    switch ((insn & 0x0fff0fff)) {
+    case 0x0e070f90:
         /* 0,c7,c0,4: Standard v6 WFI (also used in some pre-v6 cores).
          * In v7, this must NOP.
          */
+        if (IS_USER(s)) {
+            return 1;
+        }
         if (!arm_feature(env, ARM_FEATURE_V7)) {
             /* Wait for interrupt.  */
             gen_set_pc_im(s->pc);
             s->is_jmp = DISAS_WFI;
         }
         return 0;
-    }
-
-    if ((insn & 0x0fff0fff) == 0x0e070f58) {
+    case 0x0e070f58:
         /* 0,c7,c8,2: Not all pre-v6 cores implemented this WFI,
          * so this is slightly over-broad.
          */
-        if (!arm_feature(env, ARM_FEATURE_V6)) {
+        if (!IS_USER(s) && !arm_feature(env, ARM_FEATURE_V6)) {
             /* Wait for interrupt.  */
             gen_set_pc_im(s->pc);
             s->is_jmp = DISAS_WFI;
             return 0;
         }
-        /* Otherwise fall through to handle via helper function.
+        /* Otherwise continue to handle via helper function.
          * In particular, on v7 and some v6 cores this is one of
          * the VA-PA registers.
          */
+        break;
+    case 0x0e070f3d:
+        /* 0,c7,c13,1: prefetch-by-MVA in v6, NOP in v7 */
+        if (arm_feature(env, ARM_FEATURE_V6)) {
+            return IS_USER(s) ? 1 : 0;
+        }
+        break;
+    case 0x0e070f95: /* 0,c7,c5,4 : ISB */
+    case 0x0e070f9a: /* 0,c7,c10,4: DSB */
+    case 0x0e070fba: /* 0,c7,c10,5: DMB */
+        /* Barriers in both v6 and v7 */
+        if (arm_feature(env, ARM_FEATURE_V6)) {
+            return 0;
+        }
+        break;
+    default:
+        break;
+    }
+
+    if (IS_USER(s) && !cp15_user_ok(env, insn)) {
+        return 1;
     }
 
     rd = (insn >> 12) & 0xf;
-- 
1.7.1

  parent reply	other threads:[~2011-07-27 13:56 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-27 13:47 [Qemu-devel] [PULL 0/4][STABLE] ARM patch queue (for 0.15) Peter Maydell
2011-07-27 13:47 ` [Qemu-devel] [PATCH 1/4] target-arm: Mark 1136r1 as a v6K core Peter Maydell
2011-07-27 13:47 ` Peter Maydell [this message]
2011-07-27 13:47 ` [Qemu-devel] [PATCH 3/4] target-arm: Handle UNDEF and UNPREDICTABLE cases for VLDM, VSTM Peter Maydell
2011-07-27 13:47 ` [Qemu-devel] [PATCH 4/4] target-arm: UNDEF on a VCVTT/VCVTB UNPREDICTABLE to avoid TCG assert Peter Maydell

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1311774476-23321-3-git-send-email-peter.maydell@linaro.org \
    --to=peter.maydell@linaro.org \
    --cc=jmforbes@linuxtx.org \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).